ncaa_eval.cli.main module

Typer CLI application for NCAA_eval training.

ncaa_eval.cli.main.export(run_id: str = <typer.models.OptionInfo object>, season: int = <typer.models.OptionInfo object>, data_dir: Path = <typer.models.OptionInfo object>, output: Path | None = <typer.models.OptionInfo object>) → None

Export model predictions as a Kaggle submission CSV.

Generates a CSV with all pairwise matchup probabilities for men’s D1 teams in the target season, formatted for Kaggle March Machine Learning Mania competition submission.

ncaa_eval.cli.main.predict(run_id: str = <typer.models.OptionInfo object>, season: int = <typer.models.OptionInfo object>, data_dir: Path = <typer.models.OptionInfo object>, output: Path | None = <typer.models.OptionInfo object>) → None

Generate win-probability predictions for a season.

Produces a CSV with pairwise matchup probabilities (stateful models) or game-level predictions (stateless models) for the target season.

ncaa_eval.cli.main.train(model: str = <typer.models.OptionInfo object>, start_year: int = <typer.models.OptionInfo object>, end_year: int = <typer.models.OptionInfo object>, data_dir: Path = <typer.models.OptionInfo object>, output_dir: Path = <typer.models.OptionInfo object>, config: Path | None = <typer.models.OptionInfo object>) → None

Train a model on NCAA basketball data and persist run artifacts.

Validates the input year range, resolves the registered model class by name, instantiates it with optional JSON config overrides (via Pydantic validation), then delegates to run_training() with parsed CLI args.